This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS65983B: Updating from TPS65983 to TPS65983B

Part Number: TPS65983B

I'm trying to update an existing Alpine ridge single Thunderbolt port AC powered design from the TPS65983 to TPS65983B.  My new board using the original TPS65983 firmware works fine and boots up without problem, but when I create new firmware to support USB PD3 for the TPS65983B and load it onto my prototype, the board initializes and everything powers up, then immediately it shuts back down and goes back into an unconnected state.  Perhaps I have something misconfigured?  There is a new page of settings in the configuration tool that has no instructions "Tx Sopurce Capabilities Extended Data Block (0x44)"

Any help or guidance would be greatly appreciated.

Thank You,

Rob

  • I took a step back and tried to recreate the TPS65983 firmware that I already have, using the "TPS65983 Thunderbolt Application Customization 4_14", and the latest base image "TPS65983A_fw4_0004_65.bin".  I still have my original .pjt and .vif that I made using an earlier version of the tool and base image (version 4_63)

    When I create this new firmware the vif files are identical except for the time stamp and the bcdDevice_SOP which reports 0x465 instead of the original 0x463 (I change no settings in the  "TPS65983 Thunderbolt Application Customization 4_14"

    When I use this new firmware and merge with my original Alpine Ridge while leaving all settings unchanged, then program this new binary into my board using TDT, my unit becomes bricked.  When I power up it will configure the AR and the TPS65983, then wake up the board when I connect my host, and after a fraction of a second it goes back into a power down state and disconnects from the host.

    Can you connect me to the 4_63 version of the binary so that I can try to recreate that, and see if it is a problem with the  "TPS65983 Thunderbolt Application Customization 4_14" vs the "TPS65983A_fw4_0004_65.bin"

    Thank You,

    Rob

  • Hi Robert,

    How are you moving from the old version of the Application Customization Tool to the new one? I first recommend creating a new project based on your reference design. The project created should match your existing reference design. Loading an old project runs the risk of the configuration data not translating correctly into the new tool. Importing is better if required.

    I'd like to check this first before we look at older versions of the tool.

    Regards,

    Scott

  • Hi Scott,

    I made a fresh design using the TPS65983 GUI version 4.14.

    Selections -
    1) New Project
    2) Apex Creek
    3) Supports PD3 and PD2
    4) 1 PD port in system
    5) select Device 1 tab

    a) Transmit Source Capabilities
    b) Source Capabilities Settings

    i) Source Voltage Levels - set to 5V only

    Generated project file, low binary image, and VIF file (new_build.zip attached.)  Merged with my Alpine Ridge firmware and programmed into my board using Intel's TDT.

    I power down my test board and re-apply power.  The Alpine Ridge and TPS65983B load from serial flash and then waits for connection.  I plug in my host again with the Thunderbolt cable, the host recognizes a USB3.1 extensible host controller has been attached, but the board does not power up and the Alpine Ridge never asserts DG_PM_S0.

    I have also uploaded my older TPS65983A project which is working without issue in my design for comparison (TPS65983A.zip)

    Thanks,

    Robtps65983A.zipnew_build.zip

  • Hi Scott,

    In comparing a working unit and a non-working unit side by side, I noticed that a pattern of high/low data at about 12Hz on pins CC1 and CC2 on the working unit, and the non-working unit pins CC1 and CC2 are always low.

    I also connected a working unit and a non-working unit to my GRL USB PD analyzer, and got different results (see attached.)

    Thanks,

    Rob

    Working:

    Non-working:

  • Hi Robert,

    I believe you are seeing PD CC1/2 messaging in the working setup and then no activity in the non-working setup. Even if there is a configuration or setup issue, the device should show some kind of activity on CC1/2. An explanation is the non-working setup is failing to load device firmware or boot from it.  A scope trace on CC1/CC2/Vbus can help confirm this but the failure of the C2 to form a contract and check device capabilities is already a good check. 

    In your earlier post, you do mention that '83B loads from flash. This device should also support the 6.x PD3 capable firmware. Do you have access to the device I2C ports? Did anything change in the layout except the move from 83 to 83B?

    Thanks,

    Scott

  • Hi Scott,

    I do have the I2C clock and data brought out to a 4 pin header.

    The working and non-working boards are both the same revision of PCB, and they both have '83B devices on them, the only difference is that the working one is running my old firmware that I made for the 83 using 4_65 base image, and the non-working one is running the new '83B firmware from the 6_67 base image.

    Thank You,

    Rob

  • Can you do a quick check and load the PD2 (4.x series  FW / select PD2.0 only in the GUI). The newer '83B device can run either the PD2 or the PD2/3 firmware.

    Thanks,

    Scott